We have examined cardiovascular responses in humans to a wide range of abused drugs, including cocaine, marijuana, alcohol, nicotine, methylphenidate, morphine, dilaudid, and pentobarbitol. All these drugs increase heart rate and decrease vagal tone, particularly cocaine and marijuana. Vagal tone measures parasympathetic influences on the heart noninvasively by quantifying respiratory sinus arrhythmia, or heart rate variability entrained with respiration. Withdrawal of vagal inhibition produces tachycardia (i.e., increased heart rate). We have proposed that stimulation of the mesolimbic dopamine reward system by these abused drugs produces parallel activation of locomotor activity (Wise & Bozarth, 1987) and vagally-mediated tachycardia. This raises the possibility that measuring parasympathetic withdrawal and its associated tachycardia may prove useful as a simple index of a drug's abuse liability. This would require that psychoactive drugs that are not abused do not also show the same pattern (i.e., decreased vagal tone and increased heart rate). The purpose for this study was to determine whether the tachycardia sometimes produced by naloxone-precipitated withdrawal from opiates is mediated by decreased vagal tone. We hypothesized that this tachycardia is sympathetically rather than parasympathetically mediated because naloxone is not an abused drug. We administered 0.4 mg intramuscular naloxone to 12 opiate users who were classified as opiate-dependent on the basis of clinical history and toxicology. All 12 subjects exhibited signs of naloxone-precipitated withdrawal, as indicated by elevated scores on the Opiate Withdrawal Scale. We recorded heart rate and vagal tone continuously before and for 30 min after the intramuscular injection of naloxone. The cardiovascular response peaked from 11 to 16 min after the injection. We preselected the 8 subjects with the greatest heart rate increases in order to provide the strongest possible test of the hypothesis concerning vagal tone (i.e., vagal tone would not be expected to decrease if heart rate did not increase). For these 8 subjects, heart rate increased significantly (F(1.7)=20.3, p<005) approx. 7 to 8 beats/min, and vagal tone decreased slightly but nonsignificantly F(1,7)=4.4, n.s.). A lower frequency (approx. 0.10 Hz) rhythm in the heart rate variability spectrum increased nonsignificantly (F(1,7)=3.0, n.s.). This pattern of cardiovascular results was very different from that of any abused drug that we have assessed (listed above). We interpreted this pattern to indicate that, as predicted, the tachycardia from naloxone-precipitated withdrawal was sympathetically as opposed to parasympathetically mediated.